draft-merge-request
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest external data from project tickets and code diffs to generate pull request descriptions. This represents a potential surface for indirect prompt injection if the external sources contain malicious instructions. However, the skill's functionality is limited to generating Markdown text, posing minimal risk.
- Ingestion points:
SKILL.mdinstructs the agent to use 'primary source' (tickets/specs) and 'the diff'. - Boundary markers: None provided in the Markdown template to separate external content from instructions.
- Capability inventory: Limited to drafting text within the provided Markdown structure; no dangerous system capabilities or tool use listed.
- Sanitization: No sanitization steps for input data are described.
- [NO_CODE]: The skill consists exclusively of instructional Markdown and YAML metadata files. It does not contain any executable scripts (Python, JavaScript, shell) or configuration files that trigger code execution.
Audit Metadata